Rory Nathan

Identification of homogeneous regions for the purposes of regionalisation

Abstract This paper addresses the problem of identifying homogeneous subregions that can be considered to behave in a hydrologically similar fashion. The relative performance of several techniques is evaluated using the prediction of low flow characteristics in a heterogeneous group of 184 catchments located in southeastern Australia. The different approaches used are variously based on a combination of cluster analysis, multiple regression, principal component analysis, and the graphical representation of multi-dimensional data. The techniques presented allow assessment of the suitability of applying the derived regional equations to an ungauged catchment.

Detecting changes in streamflow response to changes in non-climatic catchment conditions: farm dam development in the Murray-Darling basin, Australia

Anecdotal evidence suggests that farm dams have an impact on the streamflow regime of a catchment. There is however a notable absence of suitable information on the nature and magnitude of these impacts. This work seeks to address this knowledge gap by identifying the nature of the impact of farm dams on the hydrology of catchments in the Murray-Darling Drainage Division, Australia. The 12 catchments selected for this analysis span zones of the Division with very different climatic and land use conditions. In 11 of the catchments considered there has been an increase in farm dam development over the last two to three decades. The remaining catchment, on the Jamieson River, has undergone no land use changes over the last 50 years and was selected as a control catchment for testing our approach. Statistically significant reductions in the quantity, potential streamflow response (PSR), were found for two catchments, the Yass River and Broadwater Creek. The average annual increase in farm dam capacity in these two catchments was 1.5 and 3.3% of mean annual flow, respectively. The remaining nine catchments displayed no statistically significant reductions in PSR for the entire period of simulation. Farm dam capacity information was available for five of the nine catchments. These five catchments displayed significantly smaller increases in farm dam capacity (not more than 0.3% of mean annual flow per year) than the Yass River and Broadwater Creek catchments.

Historical stream salinity trends and catchment salt balances in the Murray–Darling Basin, Australia

This paper summarizes the results from a study of historical stream salinity trends and catchment salt balances within the Murray-Darling Basin, Australia. A broad analysis of stream salinization was necessary to assist prediction of the increase in the effect and extent of dryland salinity across the basin. The sparseness of the water-quality data necessitated the development of an innovative statistical trend technique that also allowed for the high autocorrelation of the stream salinity data which was often present. Results showed the spatial distribution of stream salinization and identified regions of concern. The salinization characteristics of four distinct geographical regions were identified by providing a spatial analysis of catchment salt balances and stream salinity trends. The salinization behaviour of each region was also related to distinct physical processes. The most significant rising trends and catchment salt output/input ratios were in the zone with 500-800 mm year -1 rainfall in the southern and eastern dryland region.

Estimating low flow characteristics in ungauged catchments

This paper describes the development and derivation of a methodology for estimating low flow characteristics and yield in small ungauged rural catchments. The methodology has been applied to 184 catchments located in New South Wales and Victoria, in south-eastern Australia. A systems approach was adopted in which multivariate techniques were used to develop relationships between low flow parameters and climatic and land information data.

Development of a simplified unsaturated module for providing recharge estimates to saturated groundwater models

This paper describes the development of an integrated saturated/unsaturated catchment model of salinity processes and its application to an irrigated and dryland district in northern Victoria, Australia. The model simulates a coupled surface water and groundwater system in which a lumped conceptual model is used to model one-dimensional movement through the unsaturated zone, and a distributed model is used to simulate three-dimensional saturated groundwater flow. Movement of water through the unsaturated zone was found to be dominated by macropore flow, and accordingly the classical matrix flow approximations described by the Richards equation were not used. Instead, a conceptual model of the unsaturated zone was developed to simulate soil infiltration, recharge to the underlying groundwater system, capillary rise and crop evaporation on a daily time-step for each active cell within the model domain. Subsurface lateral flow between each vertical soil column is not permitted, though once in the saturated zone lateral flows are handled by the groundwater model. Salt movement through the profile is simulated using a mass balance approach, and separate conceptual formulations are used to generate salt export along the drainage lines. Copyright

Identifying the Separate Impact of Farm Dams and Land Use Changes on Catchment Yield

Summary This paper presents a method for determining the impact of farm dams on streamflows independent of changes in climate and other land-use. The method is comprised of two components: first, the assessment of the statistical significance of trends in streamflows independent of climatic variation, and secondly, the use of a simulation model to determine the proportion of the assessed trend which is attributable to the impact of farm dams. The method has been successfully applied to a number of catchments throughout Australia. Two case studies have been selected to illustrate its practical application, namely the Marne River catchment in South Australia and the Yass River catchment in New South Wales. The results obtained for these catchments indicate that the nature and magnitude of trends estimated by the two largely independent methods is entirely consistent with the documented changes in farm dam development and other land use changes. There would appear to be a direct correlation between the volume of farm dam development and the decrease in streamflow yield, such that for every 1 ML of farm dam development there is a corresponding decrease in streamflows of between 1 ML to 1.3 ML.

Assessing the Impact of Farm Dams on Streamflows, Part II: Regional Characterisation

Abstract This paper reports on a number of advances in the hydrologie characterisation of farm dams that are required in the assessment of farm dam impacts on streamflows. While the characteristics were derived for input to the TEDI simulation model, the generic nature of the information derived is applicable to any quantitative analysis of farm dam impacts, regardless of the method used. The use of topographic maps and aerial photography to identify the number and volume of farm dams is discussed, and this includes development of a functional relationship between farm dam surface area and volume. An essential component required for the estimation of farm dam impacts is the proportion of a dam’s volume used to satisfy demands, and this is investigated using a regional survey of landholders. Typical size distributions of farm dams used for irrigation, and for stock and domestic purposes are presented, as is information on the range of farm dam densities encountered across Victoria. Farm dam and catchment characteristics derived for 1600 Victorian catchments were used along with previously derived prediction equations to estimate the impacts of farm dams on a range of streamflow indices. The manner in which this regional information can be used to rapidly estimate the impacts of farm dams on specific catchments is also discussed and presented.

Assessing the Impact of Farm Dams on Streamflows, Part I: Development of Simulation Tools

Abstract Farm dams play an important role in Australian agriculture. In most cases the impact of an individual farm dam is relatively small, though the cumulative impact of farm dams on streamflows can be significant. A simulation model, Tool for Estimating Dam Impacts (TEDI) has been developed with the specific intention of modelling the hydrologie effect of farm dams on downstream flows. The TEDI model has a conceptually simple structure that was developed to be commensurate with the coarse level ofinformation that is generally available at the catchment scale. Given the simplicity of this model, it is desirable to investigate the salient assumptions to determine whether there are compelling scientific reasons to increase the level of modelling complexity. This paper describes the development and application of a detailed simulation model (called CHEAT) that was used to assess the efficacy of the simplifying assumptions used in TEDI. The CHEAT model was configured to exactly emulate TEDI, and enhancements were progressively introduced in order to simulate greater levels of hydrologic complexity. The impacts of different simplifying assumptions are assessed for three catchments, one in Victoria with low streamflow yields and a moderate density of farm dam development, and the other two are located in Victoria and South Australia, with high yields and high density of farm dam development. All three catchments are located in regions that typically support dryland farming. On the basis of the three catchments considered it is tentatively concluded that the combined effect of the simplifying assumptions in the TEDI model have a small influence on the results, and that it can be used a practical tool for the estimation of farm dam impacts.

A standard approach to baseflow separation using the Lyne and Hollick filter

Abstract The digital filtering approach to baseflow separation suggested by Lyne & Hollick (1979) has been widely used and is available in a number of computer packages. However, details of the approach used by different authors vary and so do the results. This means baseflow volumes and indices reported by different authors, and at different times, are difficult to compare. We propose a standard method for baseflow separation using the Lyne and Hollick digital filter. This includes reflecting the flow series at the start and end of the record to reduce “warm up” effects and the adoption of specific starting values for each filter pass.

A deterministic-empirical model of the effect of the capillary fringe on near-stream area runoff 2. Testing and application

Abstract The testing and application of HECNAR, a deterministic-empirical model that can account for the hydrologic effect of the capillary fringe on the near-stream area runoff, was carried out in several steps. The model was tested against the response of a laboratory model which resembled a near-stream saturated region of a cross-section, and the results of a field experiment. In each case, the simulated flow components were compared with the separated hydrograph components using a conservative tracer and the predictions of an unsaturated-saturated numerical model. The results were in reasonably good agreement. The model was then applied to a mine tailings site and the predictions were compared with numerical simulations, measured streamflows and the components of the stream hydrograph separated using 180 as a conservative tracer. The predictions of HECNAR and the numerical model were in good agreement, as were the trends of the simulated and the measured hydrographs. However, some discrepancies existed between the measured and the simulated flows. This study shows that HECNAR is computer efficient and has a relatively modest input requirement. Incorporation of the model into a larger-scale and more comprehensive watershed model could provide a valuable tool to quantify total flows and runoff source components to the stream, in watersheds in humid environments.